Are Indonesian professionals uncompetitive globally because schools don’t teach enough “advanced stuff” like AI? Or is there a deeper, more fundamental issue we’ve overlooked?
Dr. Hasanudin Abdurakhman argues the bottleneck isn’t content—it’s cognition. The priority isn’t to cram more advanced material into curricula, but to build strong habits of logical thinking, methodical troubleshooting, and clear communication—skills that make any tool, from spreadsheets to AI, actually useful at work.
Why this matters now: as industries automate and complexity rises, the premium shifts from tool operation to problem framing and reasoning. Without that foundation, advanced topics become expensive decorations that don’t translate to impact.
1) The Real Gap: From Skills to Structured Thinking
We’re often not short on skills like programming… but we fail to frame situations and formulate solutions using the information we have.
Graduates can calculate cylinder volume in school, yet struggle to estimate a tank’s volume in the field.
What this reveals:
- Transfer problem: Knowing formulas isn’t the same as applying them under ambiguity.
- Framing deficit: Teams jump to answers without defining the problem, constraints, or success criteria.
- Method neglect: The scientific method is treated as theory, not a daily workflow.
Real-world analogy:
- A developer who can build features but can’t isolate a production bug because they skip logging, baseline comparisons, or replication steps. The issue isn’t lack of tools—it’s lack of method.
Implication: Teaching AI or advanced math won’t fix poor framing. Advanced content relies on foundational reasoning to be useful.
2) A Workable Core: The Scientific Method at Work
Work is problem-solving. The method demands we start by documenting facts about the observed symptoms.
After collecting facts, analyze their relationships, look for causes, then formulate solutions.
A practical, repeatable loop:
- Observe facts — logs, measurements, conditions, steps taken. No opinions.
- Analyze relationships — correlate inputs to outputs, time-sequence events, isolate variables.
- Form hypotheses — list plausible causes; rank by likelihood and test cost.
- Design solutions — pick the smallest, safest test to disconfirm or confirm.
Field example:
- Machine “broken”? Before replacing parts, record error codes, environment changes, last-known-good config, and reproduce the failure under controlled steps. Then test one variable at a time.
Key mindset: resist shortcuts and “answer-first” bias—fast guesses often cause slower outcomes via rework.
3) Think Twice: Simulation, Planning, and Stepwise Scaling
In science, we simulate—everything is created twice: mentally, then physically.
Industry scales from lab to mass production through many steps. We must think before acting—and put it into a plan.
Concrete practices:
- Mental models and dry runs: Walk through the process on paper or whiteboard. Identify failure modes, resources, and checkpoints.
- Plan–Do–Check–Act rhythm: Write a minimal plan, define “stop if” criteria, and pre-commit to review points.
- Stage gates: Pilot small, collect data, adjust, then scale. This beats “just ship and hope.”
Outcome: Planning doesn’t waste time—skipping it does. Most avoidable failures are planning failures.
4) Safety, Rigor, and the Discipline of Confirmation
Think before acting to save not just performance—but lives. In Japan, ‘anzen kodo’: think (safe) before you move (act).
Operators point-and-call to confirm status—‘shisa kanko’—verbalizing checks to enforce attention and reduce error.
Translating to knowledge work:
- Use checklists for deploys, migrations, or incident response.
- Enforce two-person confirmations on high-risk steps.
- Practice “point and call” equivalents: read back commands, state the target, confirm the environment, then execute.
Result: Procedural rigor reduces mistakes, speeds onboarding, and raises baseline quality—especially under pressure.
5) Communication: Logic is the Language of Systems
Machines are logical, chemical reactions are logical, systems are logical—so communication must be logical too.
Workers often fail to assemble facts into a message their counterpart can digest.
Adopt a standard incident/report structure:
- Context: where/when/who; baseline vs change
- Facts: logs, metrics, traces, screenshots (no interpretations)
- Analysis: pattern relationships, constraints, eliminated causes
- Hypothesis: top candidate cause and rationale
- Plan: next smallest safe test; expected outcomes
- Result: what happened; updated hypothesis
Effect: Logical structure accelerates alignment, reduces back-and-forth, and teaches teams to think clearly by writing clearly.
Personal Take
The most valuable upgrade isn’t a new tool—it’s a shared operating system for thinking. A lightweight framework I recommend for teams:
- OPRA-CP (Observe → Pattern → Root cause → Action — Confirm → Publish)
- Observe: snapshot facts and conditions.
- Pattern: identify correlations, sequences, and anomalies.
- Root cause: rank hypotheses; attempt to falsify the top one first.
- Action: design the smallest safe test.
- Confirm: compare expected vs actual; adjust the model.
- Publish: document in a searchable log with a short, structured write-up.
Make it habitual by:
- Adding OPRA-CP sections to tickets and postmortems.
- Running weekly drills on small, low-risk issues to build fluency.
- Measuring “time to solid hypothesis” and “rework rate” as team KPIs.
In my experience, once teams shift from “answer-first” to “method-first,” velocity improves—not because they move faster, but because they redo less.
Conclusion
The main problem with Indonesia’s human capital isn’t missing AI modules—it’s missing mental models. When teams observe facts, analyze relationships, form hypotheses, plan and simulate, and communicate logically, their existing skills suddenly become leverage.
Before rushing to advanced content, strengthen the foundation: think first, then act.
Question to reflect on: What is one routine in your workflow you can “simulate twice” starting this week—once on paper, once in production—and what checklist would make it safer?
Watch the full talk for nuanced examples and context at Malaka's Masalah Utama SDM Indonesia.